It is widely recognized that humans in modern society do not get the sleep they need. This chronic sleep debt significantly degrades productivity, increases accident rates and has been linked to poor health and increased risk of death. Humans are exposed to a light dark (LD) cycle created by natural and artificial light alternating with sleep. Sleep duration determines the length of dark in the daily LD cycle. However, the effects of repeated short dark periods (short nights) on the human circadian system remain unknown. It has been proposed that mammals have a complex circadian pacemaker, consisting of an evening oscillator, that regulates the onset of melatonin secretion, and a morning oscillator, that regulates the offset of melatonin secretion. In short summer nights (longer photoperiods), the phase relationship between the evening and morning oscillators is decreased, which shortens the duration of the nocturnal melatonin profile, in turn regulating appropriate seasonal changes. Other animal work shows that the range and amplitude of the light phase response curve is decreased in short nights. We propose to investigate the effects of short and long nights on the human circadian system, by keeping subjects in bed, in the dark, for either 6 or 9 hours per night. We will determine if the duration of the melatonin rhythm in dim light compresses after a history of short nights, and expands after a history of long nights, thus revealing that the circadian pacemaker has a memory of the previous night length. We will also determine if a history of short nights reduces the phase shifting response to light. Our goal is to determine the extent to which the chronic sleep debt in modern society is altering the functioning of the human circadian system. This work has important implications for the sleep deprived general population and for jet travelers and night shift workers who must phase shift their circadian rhythms in order to adjust to phase shifts in their light/dark cycle.